Fig. 1: Integration of global cortical gradients with cortical arealization.

A Cortex-wide connectomes were constructed from microstructural profile covariance (MPC)²³, structural connectivity (SC)^59,61, and functional connectivity (FC)^22,60 at a vertex-level based on repeated 7 T MRI. We applied non-linear dimensionality reduction techniques¹²⁹ to each connectome and aggregated the first five eigenvectors/gradients. B Probabilistic area definitions were derived from the Julich-Brain atlas¹⁹, a post-mortem cytoarchitectonic atlas based on the mapping of areas of ten postmortem brains, and their superimposition in MNI space. Please note that this probabilistic atlas does not cover the entire cortex. C We averaged vertex-wise gradients in each of the 228 areas, producing area-specific multimodal gradient profiles. These gradient profiles were reordered according to their principal component to assess inter-areal similarity. Left panel: the reordered gradient profiles located in the middle and at the two ends of the main axis were visualized in spider plots. Source data are provided as a Source Data file. Middle panel: the reordered multimodal gradient profiles. Right panel: the first principal component from the original multimodal gradient profiles. Abbreviation: PCA principal component analysis, WM white matter.